JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY VOL. 69, NO.

3, 2017

ª 2017 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION ISSN 0735-1097/$36.00

PUBLISHED BY ELSEVIER http://dx.doi.org/10.1016/j.jacc.2016.10.022

Percutaneous Mechanical Circulatory

Support Versus Intra-Aortic Balloon
Pump in Cardiogenic Shock After
Acute Myocardial Infarction
Dagmar M. Ouweneel, MSC,a Erlend Eriksen, MD,b Krischan D. Sjauw, MD, PHD,a Ivo M. van Dongen, MD,a
Alexander Hirsch, MD, PHD,a Erik J.S. Packer, MD,b M. Marije Vis, MD, PHD,a Joanna J. Wykrzykowska, MD, PHD,a
Karel T. Koch, MD, PHD,a Jan Baan, MD, PHD,a Robbert J. de Winter, MD, PHD,a Jan J. Piek, MD, PHD,a
Wim K. Lagrand, MD, PHD,c Bas A.J.M. de Mol, MD, PHD,a Jan G.P. Tijssen, PHD,a José P.S. Henriques, MD, PHDa

ABSTRACT

BACKGROUND Despite advances in treatment, mortality in acute myocardial infarction (AMI) complicated by
cardiogenic shock (CS) remains high. Short-term mechanical circulatory support devices acutely improve hemodynamic
conditions.

OBJECTIVES The aim of this study was to determine whether a new percutaneous mechanical circulatory support
(pMCS) device (Impella CP, Abiomed, Danvers, Massachusetts) decreases 30-day mortality when compared with an intra-
aortic balloon pump (IABP) in patients with severe shock complicating AMI.

METHODS In a randomized, prospective, open-label, multicenter trial, 48 patients with severe CS complicating AMI
were assigned to pMCS (n ¼ 24) or IABP (n ¼ 24). Severe CS was defined as systolic blood pressure <90 mm Hg or the
need for inotropic or vasoactive medication and the requirement for mechanical ventilation. The primary endpoint was
30-day all-cause mortality.

RESULTS At 30 days, mortality in patients treated with either IABP or pMCS was similar (50% and 46%, respectively;
hazard ratio with pMCS: 0.96; 95% confidence interval: 0.42 to 2.18; p ¼ 0.92). At 6 months, mortality rates for both
pMCS and IABP were 50% (hazard ratio: 1.04; 95% confidence interval: 0.47 to 2.32; p ¼ 0.923).

CONCLUSIONS In this explorative randomized controlled trial involving mechanically ventilated patients with CS after
AMI, routine treatment with pMCS was not associated with reduced 30-day mortality compared with IABP. (IMPRESS in
Severe Shock; NTR3450) (J Am Coll Cardiol 2017;69:278–87) © 2017 by the American College of Cardiology Foundation.

D
high
espite advances in treatment, mortality in
acute myocardial infarction (AMI) compli-
cated by cardiogenic shock (CS) remains
(1–4). Short-term mechanical
support devices can be deployed to support the
endangered circulation. Intra-aortic balloon counter-
pulsation (IABP) has been the most widely used
circulatory
mechanical circulatory support device for decades
(5). A meta-analysis of smaller-sized studies and a
large randomized controlled trial did not show a
beneficial effect of IABP in the setting of CS after
AMI (4,6,7). Today, IABP usage has a Class IIb
recommendation in American guidelines and a Class
III recommendation in European guidelines (8–11).

Listen to this manuscript’s

audio summary by
JACC Editor-in-Chief
Dr. Valentin Fuster.
From the aAMC Heart Center, Academic Medical Center–University of Amsterdam, Amsterdam, the Netherlands; bDepartment of
Heart Disease, Haukeland University Hospital, Bergen, Norway; and the cDepartment of Intensive Care Medicine, Academic
Medical Center–University of Amsterdam, Amsterdam, the Netherlands. The trial was funded by the Academic Medical Center,
Amsterdam. The Academic Medical Center has received research grants and speaker honoraria from Abiomed Inc. Dr. Piek has
served on the medical advisory board of Abbott Vascular; and has served as a consultant for Philips Volcano. All other authors
have reported that they have no relationships relevant to the contents of this paper to disclose.

Manuscript received October 10, 2016; accepted October 18, 2016.

JACC VOL. 69, NO. 3, 2017
JANUARY 24, 2017:278–87
The lack of efficacy of the IABP is likely to be, at
least partly, the reason for the observed increased
usage of more potent mechanical circulatory devices
(5,12).
The percutaneous Impella platform (Abiomed,
Danvers, Massachusetts) consists of the Impella 2.5
(maximum output 2.5 l/min) and Impella
(maximum output around 3.7 l/min). It has been
shown that Impella support in the acute situation is
feasible and provides greater hemodynamic support
when compared with IABP (13–16). However, neither of
the 2 small randomized trials in patients with AMI had
enough power to show differences in clinical out-
comes, and 1 was prematurely stopped (15,16). The
IMPRESS in Severe Shock (IMPella versus IABP Re-
duces mortality in STEMI patients treated with pri-
mary PCI in Severe cardiogenic SHOCK) trial is an
exploratory assessment of mortality and other safety
outcomes comparing percutaneous mechanical circu-
latory support (pMCS) by the Impella CP with IABP in
mechanically ventilated patients with CS in AMI.
SEE PAGE 288
METHODS
STUDY DESIGN. The Academic Medical Center in
Amsterdam designed and sponsored this multicenter,
open-label, and randomized trial. Trial administra-
tion, data management, and statistical analysis were
performed by the sponsor. The executive committee
had unrestricted access to the data, and the authors
analyzed the data and prepared the paper. The trial
design was approved by the ethics committee at each
participating center. The ethics committee waived
the requirements for written informed consent before
randomization to prevent treatment delay in patients
who were in imminent danger of death. The
requirement for obtaining informed consent to use
the data varied depending on local ethical re-
quirements. Informed consent was obtained from the
legal representative without any undue delay. Alter-
natively, informed consent was obtained after re-
covery (and therefore, no informed consent was
obtained in patients who died). An independent data
and safety monitoring board and the ethics commit-
tees reviewed the interim results after each 10
included patients. During the inclusion period of the
trial, the European Society of Cardiology guidelines
for routine use of the IABP changed from Class II (may
be considered) to Class III (not recommended) (10).
The ethics committees were notified of this change
and approved continuation of the trial. The study was
conducted in accordance with the provisions of the
Ouweneel et al. 279
Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock
Declaration of Helsinki as amended in Edin- ABBREVIATIONS
burgh, Scotland (17). AND ACRONYMS
PATIENTS. Patients were eligible for the trial AMI = acute myocardial
if they presented with an AMI with ST- infarction
segment elevation complicated by severe CS CS = cardiogenic shock
in the setting of immediate percutaneous IABP = intra-aortic balloon
CP coronary intervention (PCI). Severe CS was pump
defined as a systolic blood pressure <90 LVAD = left ventricular assist
mm Hg for longer than 30 min or the need for device
inotropes or vasopressors to maintain a sys- PCI = percutaneous cardiac
tolic blood pressure >90 mm Hg. To select a
intervention
patient population in even worse condition, pMCS = percutaneous
mechanical circulatory support
patients only qualified if they were mechan-
ROSC = return of spontaneous
ically ventilated before randomization. circulation
Exclusion criteria were: severe aorto-iliac
arterial disease impeding placement of either IABP
or pMCS, known severe cardiac aortic valvular dis-
ease, serious known concomitant disease with a life
expectancy of <1 year, known participation in this
study or any other trial within the previous 30 days,
or coronary artery bypass grafting within the pre-
ceding week.
TREATMENT. Eligible patients were assigned to
treatment with pMCS by Impella CP with IABP (con-
trol group). Randomization was performed in a 1:1
ratio using an internet-based application. The
moment of initiation of pMCS or IABP (before, during,
or immediately after the PCI) was at the discretion of
the treating physician. To achieve equal initiation of
therapy for both groups, timing of randomization was
equal to pMCS or IABP placement: immediately
before, during, or after PCI.
All patients underwent primary PCI. In multi-
vessel disease, the mode of revascularization (im-
mediate or staged PCI of the nonculprit lesions) was
left to the discretion of the operator. Duration of
mechanical support was left to the discretion of the
treating physician, and IABP or the pMCS device was
extracted in accordance with daily clinical routine.
Weaning was achieved by reduction of the trigger
ratio (IABP) or amount of support (pMCS). Per pro-
tocol, crossover was not allowed; however, it did
occur.
OUTCOMES. The primary study endpoint was 30-day
all-cause mortality. The secondary endpoint was
6-month mortality. Descriptive endpoints included
duration of mechanical ventilation; the need for and
duration of inotropic and vasopressor therapy; renal
replacement therapy; length of hospital stay; the
amount of blood products needed; additional treat-
ments, such as ICD placement and the need for sur-
gical left ventricular assist device (LVAD) placement
280 Ouweneel et al. JACC VOL. 69, NO. 3, 2017

Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock JANUARY 24, 2017:278–87

F I G U R E 1 Flow Diagram

Academic Medical Center, Amsterdam, the Netherlands Haukeland University Hospital, Bergen, Norway
screening between May 24, 2012 and September 15, 2015 screening between February 11, 2014 and September 15, 2015

75 eligible mechanically ventilated cardiogenic shock patients

with STEMI undergoing primary PCI

27 patients not enrolled:

- severe peripheral vessel disease (n=2)
- left ventricular thrombus (n=1)
- ongoing resuscitation (n=9)
- operator’s preference for specific assist device
- IABP (n=7)
- pMCS (n=8)

48 randomized patients

24 randomized to IABP: 24 randomized to pMCS:

- 2 upgrade to ImpelIa 5.0 - 1 randomized while already on IABP support
- 1 upgrade to pMCS - 1 upgrade to Impella 5.0
- 1 received IABP
- 1 received no device

- 1 lost to follow-up after 31 days - none lost to follow-up

Flow chart of the screening, randomization, and follow-up of the patients.

or heart transplantation; the occurrence of stroke, difference in mortality between the treatment
myocardial reinfarction, repeat PCI, coronary artery groups. Therefore, adaptation of the sample size was
bypass grafting, major vascular complications, major not meaningful, and the Executive Committee
bleeding, or hemolysis requiring extraction of the decided to complete the study with 48 patients as an
IABP or pMCS; device failure requiring extraction of exploratory safety study.
the pMCS or IABP; and rehospitalization. Definitions All data were analyzed according to the intention-
can be found in the supplementary file. An indepen- to-treat principle. In addition, a per-protocol anal-
dent clinical event committee adjudicated the events. ysis of the primary endpoint was performed.
Imaging parameters were assessed by independent Cumulative mortality throughout the first 6 months
local core laboratories that were blinded to the other following randomization was characterized with the
trial data and randomization outcome (Online use of Kaplan-Meier plots, with the log-rank test used
Appendix). for the comparison between the 2 groups. Descriptive
STATISTICAL ANALYSIS. On the basis of previous endpoints and clinical course variables were not sta-
studies and our experience that survival is <10% in tistically tested because they are highly influenced by
patients with severe shock, we assumed that treat- the number of deceased patients in both groups.
ment with pMCS would decrease the absolute 30-day Additional comparisons were made according to vital
mortality rate from 95% to 60%. On the basis of this status at 30 days. Differences were assessed with the
assumption, a trial with 24 patients in each group Fisher exact test or the chi-square test for binary
would achieve 80% power, with a 2-sided alpha endpoints and a Mann-Whitney U test for quantita-
of 5%. The protocol allowed for a sample-size tive endpoints.
re-estimation after inclusion of 32 patients. At the A post hoc subgroup analysis was performed in
interim analysis, mortality in the control group was subgroups defined according to age (<75 or >75 years),
much lower than anticipated, and there was no sex, time to return of spontaneous circulation (ROSC)
JACC VOL. 69, NO. 3, 2017 Ouweneel et al. 281
JANUARY 24, 2017:278–87 Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock

(>20 or <20 min), lactate level >7.5 or <7.5 mmol/l,

T A B L E 1 Baseline Characteristics
TIMI (Thrombolysis In Myocardial Infarction) flow
post-PCI, systolic blood pressure before IABP or pMCS pMCS IABP
(n ¼ 24) (n ¼ 24)
placement (>80 or <80 mm Hg), and the presence or
Age, yrs 58  9 59  11
absence of traumatic injuries on admission.
Male 18/24 (75) 20/24 (83)
Body mass index, kg/m2 25 (23–26) 26 (25–27)
RESULTS
Cardiovascular risk factors
Current smoking 11/18 (61) 6/19 (32)
PATIENTS. Between June 1, 2012, and September 15, Hypertension 4/20 (20) 6/21 (29)
2015, a total of 48 patients were randomly assigned to Hypercholesterolemia 4/20 (20) 5/21 (24)
either pMCS therapy (n ¼ 24) or IABP (n ¼ 24) Diabetes mellitus 2/22 (9) 3/23 (13)

(Figure 1). The baseline characteristics of the 2 groups Prior myocardial infarction 1/22 (5) 1/23 (4)
Prior stroke 0/22 (0) 1/23 (4)
were well balanced (Table 1). The mean age was
Known peripheral arterial disease 2/23 (9) 0/23 (0)
58 years, 79% were male, all patients were mechani-
Prior PCI or CABG 1/22 (5) 0/23 (0)
cally ventilated, 96% of the patients received cate- Hemodynamic variables before randomization
cholamines, and 92% had had a cardiac arrest before Heart rate, beats/min 81  21 83  28
randomization. Mean arterial pressure, mm Hg 66  15 66  15
Systolic blood pressure, mm Hg 81  17 84  19
TREATMENT. Randomization and placement of the
Diastolic blood pressure, mm Hg 58  22 57  13
pMCS or IABP took place after revascularization
Medical therapy before randomization
except for 8 patients in whom IABP or the pMCS was
Catecholamines or inotropes 24/24 (100) 22/24 (92)
initiated before revascularization (3 in the IABP group Mechanical ventilation 24/24 (100) 24/24 (100)
and 5 in the pMCS group) (Table 2). The infarct- Cardiac arrest before randomization 24/24 (100) 20/24 (83)
related artery was the left anterior descending (LAD) Witnessed arrest 22/24 (92) 17/20 (85)
in the majority of the patients (65%) and 98% of the First rhythm VT/VF 22/24 (92) 17/20 (85)

patients were treated with stent placement. The me- Time till return of spontaneous 21 (15–46) 27 (15–52)
circulation, min
dian duration of circulatory support was 48 h (IABP)
Traumatic injuries at admission 5/24 (21) 2/24 (8)
and 49 h (pMCS). All patients were treated with cat- Blood values on admission*
echolamines during admission to the intensive care Lactate, mmol/l 7.5  3.2 8.9  6.6
unit, 31% received renal replacement therapy, and Hemoglobin, mmol/l 8.6  1.2 8.6  1.2
75% were treated with therapeutic hypothermia Creatinine, mg/dl 96  29 102  22

(Table 3). Glucose, mmol/l 16.2  4.7 14.1  5.3

Arterial pH 7.14  0.14 7.17  0.17
Of the patients in the IABP group, 1 patient sub-
Baseline echocardiography†
sequently received pMCS and was transferred to
Estimated left ventricular ejection fraction
another hospital for treatment with extracorporeal <20% 5/22 (23) 8/18 (44)
life support oxygenation. Two patients received an 20%–40% 10/22 (46) 6/18 (33)
alternative device, the Impella 5.0 (Abiomed, Aachen, >40% 7/22 (32) 4/18 (22)
Germany), after the IABP treatment, and 1 of them
Values are mean  SD, n/N (%), or median (25th to 75th percentile). *Values are present for the following
received subsequent extracorporeal life support and number of patients: lactate (21 IABP and pMCS), hemoglobin (22 IABP and 21 pMCS), creatinine (23 IABP and 23
an LVAD at another hospital. Of the patients treated pMCS), glucose (23 IABP and 20 pMCS), and pH (16 IABP and 20 pMCS). †First echocardiogram made during
admission during the first day. In 20 patients, the echocardiography was done before pMCS or IABP placement,
with the pMCS, 1 patient subsequently received the and in 21 patients, after pMCS or IABP placement (within 24 h).
Impella 5.0. One patient was already on IABP support CABG ¼ coronary artery bypass graft; IABP ¼ intra-aortic balloon pump; PCI ¼ percutaneous coronary inter-
vention; pMCS ¼ percutaneous mechanical circulatory support; VF ¼ ventricular fibrillation; VT ¼ ventricular
before randomization (inserted before the start of the tachycardia.
primary PCI) and was randomized after the PCI to
pMCS treatment. Formally, this patient constitutes a
protocol violation, as IABP therapy before randomi- (HR: 1.04; 95% CI: 0.47 to 2.32; p ¼ 0.92). Only minor
zation was an exclusion criterion. One patient did not differences were found in an analysis restricted to the
receive pMCS as the patient showed signs of recovery per-protocol population from which 3 patients
after randomization to receive device therapy. treated with pMCS were excluded (Online Appendix).
OUTCOMES. At 30 days, mortality was similar in pa- The Kaplan-Meier estimates for 6-month mortality in
tients treated with IABP and pMCS therapy: 50% and the per-protocol population were 52% in the IABP
46%, respectively (hazard ratio [HR] with pMCS group and 48% in the pMCS group (HR with pMCS:
therapy: 0.96; 95% confidence interval [CI]: 0.42 to 0.95; 95% CI: 0.41 to 2.21; p ¼ 0.91).
2.18; p ¼ 0.92) (Table 4, Central Illustration). At 6 The primary cause of death at 6 months was brain
months, the mortality rate was 50% in both groups damage (46% of the deceased patients; 6 of 12 in the
282 Ouweneel et al. JACC VOL. 69, NO. 3, 2017

Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock JANUARY 24, 2017:278–87

31 days, and 1 was bedbound due to multiple scle-

T A B L E 2 Procedural Characteristics
rosis. Left ventricular ejection fraction after 2.5
pMCS IABP months (median 191 days) was 46  11% in the pMCS
(n ¼ 24) (n ¼ 24)
group and 49  9% in the IABP group.
Moment of device placement
Subgroup analysis showed no significant interac-
Device placement before 5/24 (21) 3/24 (13)
revascularization tion in 30-day mortality between the IABP and pMCS-
Device placement after 19/24 (80) 21/24 (88) treated patients with respect to age, sex, ROSC times,
revascularization
lactate levels on admission, moment of IABP or pMCS
Infarct-related artery
Left main 1/24 (4) 2/24 (8)
placement, systolic blood pressure before device
Left anterior descending 16/24 (67) 15/24 (63) placement, and traumatic injuries on admission
Left circumflex 6/24 (25) 3/24 (13) (Online Appendix, Online Table 1).
Right coronary artery 1/24 (4) 4/24 (17) When analyzing the combined study population,
Multivessel disease* 15/24 (63) 21/24 (88) lower 30-day mortality rates were seen in patients
Immediate PCI of nonculprit lesion 3/15 (20) 4/21 (19)
who had ROSC in <20 min (19% vs. 70%; HR: 5.50;
Stent placement 23/24 (96) 24/24 (100)
95% CI: 1.82 to 16.58; p ¼ 0.001) and patients with
Drug-eluting stent 22/23 (96) 22/24 (92)
lactate level on admission lower than 7.5 mmol/l (29%
Bare-metal stent 1/23 (4) 2/24 (8)
TIMI flow pre-PCI
vs. 60%; HR: 3.09; 95% CI: 1.09 to 8.74; p ¼ 0.04)
0 or 1 20/24 (83) 20/24 (83) (Online Appendix, Online Table 2). A trend toward
2 or 3 4/24 (17) 4/24 (17) lower 30-day mortality was observed if therapy with
TIMI flow post-PCI pMCS or IABP was initiated before the primary PCI
0 or 1 1/24 (4) 0/24 (0) (25% vs. 53%; p ¼ 0.16) and in patients who did not
2 or 3 23/24 (96) 24/24 (100)
have traumatic injuries (44% vs. 71%; HR: 1.88; 95%
SYNTAX score pre-PCI 23.2  8.7 28.2  10.6
CI: 0.70 to 5.07; p ¼ 0.18) (Online Appendix). Trends
Values are n/N (%) or mean  SD. *>50% stenosis in nonculprit vessel. in lactate and creatinine levels and inotrope and va-
TIMI ¼ Thrombolysis In Myocardial Infarction; other abbreviations as in Table 1. sopressors usage can be seen in Online Figures 1 to 4.
Also, characteristics of the survivors versus the non-
survivors and more extensive cardiac function pa-
IABP group and 5 of 12 in the pMCS group). Death due rameters are described in Online Tables 3 and 4.
to refractory CS occurred in 29% of the deceased pa-
tients (3 of 12 in the IABP vs. 4 of 12 in the pMCS DISCUSSION
therapy group).
In each group, 1 patient experienced an ischemic This is the first randomized trial to compare Impella
stroke during support. There was 1 major vascular CP with the IABP in mechanically ventilated patients
complication in the pMCS group, a retroperitoneal with CS complicating AMI. pMCS support was not
bleeding after pMCS insertion (the patient had a associated with lower 30-day or 6-month mortality
calcified and stented vascular trajectory, but femoro- when compared with IABP support. Although this
iliac angiography seemed compatible with pMCS trial included only 48 patients, it is thus far the
insertion, see Online Appendix for event specifica- largest trial to randomly compare pMCS and IABP,
tions). There were more bleeding events during and it is the only trial to use the Impella CP device.
admission in the pMCS therapy group than in the To date, only a few randomized controlled trials
IABP group (8 vs. 2, of which 3 and 1, respectively, have studied mechanical circulatory support in CS,
were adjudicated as IABP or pMCS related). There highlighting the logistical and ethical challenges in
were 2 patients in whom the presence of hemolysis conducting trials in these patients. In the setting of CS,
influenced the decision to stop the pMCS support (in 2 small trials have been performed with the Impella 2.5
one patient, the pMCS support was stopped due to pMCS, both using IABP therapy as the control therapy.
hemolysis in combination with an improved ejection The ISAR-SHOCK (Efficacy Study of LV Assist Device to
fraction; and in the other patient, the pMCS Treat Patients With Cardiogenic Shock) trial random-
was removed after the decision to withhold further ized 26 patients between IABP and the Impella 2.5 in
therapy due to multiorgan failure, recurrent ven- the setting of CS complicating AMI. The primary
tricular arrhythmia, hemolysis, and hemodynamic endpoint was the difference in cardiac index after 30
instability). min of support, and the trial showed a higher cardiac
Follow-up echocardiography was performed and index in patients treated with Impella than with IABP.
collected in all survivors except for 3 patients: 1 Overall mortality was 46% in both groups (15). The
received a surgical LVAD, 1 was lost to follow-up after IMPRESS in STEMI trial randomized between the IABP
JACC VOL. 69, NO. 3, 2017 Ouweneel et al. 283
JANUARY 24, 2017:278–87 Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock

and Impella 2.5 in patients with cardiogenic pre-

T A B L E 3 Clinical Course During Admission
shock. This study was powered for a difference in
left ventricular function. However, this trial was pMCS (n ¼ 24) IABP (n ¼ 24)

stopped prematurely due to a lack of enrollment after Mechanical circulatory support

Duration of support (h)* 49 (28–76) 48 (24–77)
21 patients had been enrolled (16).
Crossover or upgrading to device with 1/24 (4.2) 3/24 (12.5)
In the present trial, we included mechanically more support†
ventilated patients with CS. Although the decision to Other support before randomization‡ 1/24 (4.2) 0/24 (0)
start mechanical ventilation may be arbitrary and the Mechanical ventilation
moment of initiation may differ between physicians, Patients treated 24/24 (100) 24/24 (100)

it is a marker for worse clinical condition. We have Duration (days since device placement) 4 (3–9) 4 (3–10)
Catecholamines
chosen to use this criterion because it is easy to apply,
Patients treated 24/24 (100) 24/24 (100)
is readily available, and does not require blood sam-
Duration (days) 3 (2–6) 3 (2–5)
ple analysis or additional Swan-Ganz cardiac output Inotropic therapy (dobutamine)
measurements. Those inclusion criteria resulted in Patients treated 6/24 (25) 9/24 (38)
inclusion of patients with high lactate and low pH Duration (days) 0 (0–1) 0 (0–2)
levels on admission, and all patients received cate- Renal replacement therapy

cholamines before randomization. Although we did Patients treated 8/24 (33) 7/24 (29)
Duration (days) 17 (5–29) 7 (2–9)
not aim to include resuscitated patients, the inclusion
Therapeutic hypothermia
criteria resulted in 92% of enrolled patients having a
Patients treated 19/24 (79) 17/24 (71)
cardiac arrest prior to randomization. In addition,
Premature ending of therapeutic hypothermia 3/19 (16) 1/17 (6)
almost one-half (48%) of the patients had time to Blood products during admission§
ROSC longer than 20 min. Traumatic injuries due to Any blood products during admission 11/24 (46) 8/24 (33)
cardiac arrest were frequently present (15%). These Packed red blood cells
criteria identified a unique patient population that is Patients treated 11/24 (46) 8/24 (33)

usually excluded from randomized CS clinical trials
and resulted in a patient population with a high 30-
day mortality rate of 48%. This is higher than in the
most recently reported randomized trial on CS (IABP-
SHOCK II [Intra-Aortic Balloon Counterpulsation in
Acute Myocardial Infarction Complicated by Cardio-
genic Shock] trial), which reports a mortality of 40%
in patients randomized between IABP support and
conventional therapy (n ¼ 598) (4). Two
previous studies compared IABP and TandemHeart
(CardiacAssist Inc., Pittsburgh, Pennsylvania) in CS,
with 30-day mortality rates of 44% (n ¼ 41) (18) and
42% (n ¼ 33) (19). Neither trial observed any differ-
ence in mortality between the patients treated with
TandemHeart or IABP. A registry reporting on Impella
2.5 versus IABP in the setting of post-cardiac arrest
shock reports mortality rates of 77% in patients
treated with the device and 79% in patients treated
with IABP (20). Two multicenter registries including
patients with CS complicating AMI supported with a
pMCS showed mortality at discharge of 49.3% (n ¼ 154)
and 30-day mortality of 64.2% (n ¼ 120) (21,22).
A recent USpella registry analysis submitted to the
U.S. Food and Drug Administration for the Impella
pre-market approval for use in CS demonstrated a
marked difference between patients who were likely
to be included in randomized shock trials versus those
who were not—the latter of whom resemble the pop-
ulation studied in the present trial (23). A considerable
proportion of patients died due to anoxic brain
Number of units administered 6 (3–13) 3 (1–5)
Fresh frozen plasma 3/24 (13) 0/24 (0)
Platelets 4/24 (17) 1/24 (4)
Placement of implantable cardioverter-defibrillator 2/24 (8) 1/24 (4)
Length of stay (days)
Intensive care unit 7 (3–16) 7 (4–10)
Hospital 16 (3–26) 10 (6–24)
Values are median (25th to 75th percentile) or n/N (%). *Sum of support duration of all given support devices,
including upgrades. †One patient was upgraded from IABP to pMCS and transferred to another hospital to receive
extracorporeal life support; 1 patient received pMCS and was upgraded to Impella 5.0; 1 patient was upgraded
from IABP to Impella 5.0; and 1 patient was upgraded from IABP to Impella 5.0 and transferred to another
hospital to receive extracorporeal life support and surgical LVAD. ‡One patient was already on IABP support
before randomization and was randomized to pMCS support. §Only blood products in the hospital of randomi-
zation are taken into account.
LVAD ¼ left ventricular assist device; other abbreviations as in Table 1.
damage (46%), compared with refractory CS or mul-
tiorgan failure (29%), or for other reasons (25%). This
high rate of neurologically deceased patients is likely
to be the result of the high percentage of resuscitated
patients and longer times to ROSC. Nevertheless, our
study resembles a real-life cohort in daily clinical
practice of patients with CS complicating ST-segment
elevation myocardial infarction.
In our study, bleeding occurred more often in the
pMCS-treated patients than in the IABP-treated pa-
tients. During mechanical support, patients receive
heparin in addition to standard dual antiplatelet
therapy after PCI (aspirin and a P2Y 12 receptor
blocker), which makes the occurrence of bleeding
more likely, especially in patients with additional
traumatic injuries on admission. Higher rates of
284 Ouweneel et al. JACC VOL. 69, NO. 3, 2017

Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock JANUARY 24, 2017:278–87

The IABP-SHOCK II trial reports 20.7% bleeding in the

T A B L E 4 Clinical and Functional Outcomes
IABP patients (and 20.8% in the control group). This
pMCS IABP Hazard Ratio With is higher than the 8.2% bleeding in our IABP group.
(n ¼ 24) (n ¼ 24) p Value pMCS (95% CI)
Although discouraged, some crossovers and up-
Mortality*
grades to other mechanical support therapy did take
30-day all-cause mortality 11 (46) 12 (50) 0.92 0.96 (0.42–2.18)
6-month all-cause mortality 12 (50) 12 (50) 0.92 1.04 (0.47–2.32) place: 3 in the IABP group and 1 in the pMCS group.
Clinical outcomes at 6 months Crossover or upgrading was solely at the discretion of
Cause of death the investigator. There was a trend toward more
Refractory cardiogenic shock 4 (17) 3 (13) upgrading/crossover in the IABP group.
Post-anoxic neurological death 5 (21) 6 (25) Upon initiation of our study, IABP therapy was still
Other reason 3 (13) 3 (13)
recommended in the guidelines for CS, but was
Stroke 1 (4) 1 (4)
downgraded to a Class III recommendation in the
Hemorrhagic stroke 0 (0) 0 (0)
Ischemic stroke 1 (4) 1 (4)
European guidelines and Class II in the American
Major vascular complication 1 (4) 0 (0) guidelines during the inclusion period of the study.
Major bleeding 8 (33) 2 (8) After consultation with the institutional review board
Device-related bleeding 3 (13) 1 (4) and in the light of the severity of clinical condition
Retroperitoneal 1 (4) 0 (0) with higher mortality rates than in the IABP-SHOCK II
IABP/Impella puncture site 2 (8) 1 (4)
study, the control therapy remained unchanged. In
Nondevice-related bleeding 5 (21) 1 (4)
addition, after the interim analysis it was clear the
Gastro-intestinal bleeding 0 (0) 1 (4)
study was underpowered to show a difference in
Bleeding at other puncture site 1 (4) 0 (0)
Other location 4 (17) 0 (0) mortality at 30 days, and the executive committee
Hemolysis requiring extraction 2 (8) 0 (0) allowed it to proceed for exploratory purposes.
of the device Although not adequately powered, our trial sug-
Device failure requiring extraction 0 (0) 0 (0)
gests that in patients with CS without selection on
Surgical LVAD placement 0 (0) 1 (4)
age, ROSC times, and pre-procedural traumatic in-
Heart transplantation 0 (0) 0 (0)
Other surgery 2 (8) 0 (0)
juries, no clear signal of superior outcome was
Myocardial (re)infarction 1 (4) 2 (8) observed in patients with pMCS support when
Repeat PCI 0 (0) 3 (13) compared with the IABP.
CABG 0 (0) 1 (4) There may be several reasons why the pMCS-
Rehospitalization 5 (21) 1 (4) treated patients did not show improved mortality
Cardiac 2 (8) 0 (0)
rates. A possible explanation is the unselective nature
Noncardiac 3 (13) 1 (4)
of the patients included in the study. In our study,
Echocardiography at 6 months†
92% of patients had resuscitated cardiac arrest, which
Left ventricular dimensions and 12 9‡
systolic function (n) implies a prevalence of post-anoxic neurological
Ejection fraction (%) 46  11 49  9 damage present at the moment of randomization.
End-diastolic volume (ml) 122  41 120  33 Any kind of mechanical circulatory support may be of
End-systolic volume (ml) 65  31 61  21
limited clinical utility in these patients. Another

Values are frequencies (%) or mean  SD, unless otherwise indicated. Additional information about events can be
found in the Online Appendix. *Mortality is shown as Kaplan-Meier estimates. †First available echocardiogram
after 2 months. Median follow-up time is 191 (176 to 297) days. ‡One patient with surgical LVAD, 1 patient lost to
follow-up, and 1 patient bedridden due to multiple sclerosis.
CI ¼ confidence interval; other abbreviations as in Tables 1 and 3.
bleeding in pMCS-treated patients compared with
IABP-treated patients were also described in a
registry comparing Impella 2.5 and IABP in a post-
cardiac arrest population (n ¼ 78), which found
severe bleeding in 26% of the device patients versus
6% of IABP patients (20). Two large multicenter
Impella 2.5 registries describe the rates of bleeding
requiring transfusion of 24.2% and 17.5% and the
rates of hemolysis as 7.5% and 10.3% (21,22). These
complication rates are comparable to the pMCS in
our study (33.3% bleeding and 8.3% hemolysis).
explanation might be that CS after AMI is not only a
matter of low cardiac output. The shock syndrome
also comprises an irreversible damage due to dimin-
ished organ perfusion and inflammatory responses.
Hence, providing mechanical hemodynamic support
may not be enough to reverse the damage that has
already occurred. Although the Impella CP can pro-
vide up to 3.5 l/min of forward flow, it might still be
insufficient to reverse severe CS with advanced end
organ failure, especially as in clinical practice, long-
term Impella CP support achieves <3.5 l/min hemo-
dynamic support. In this trial, the main rationale
for using Impella CP instead of a device that can
provide even more hemodynamic support (e.g.,
Impella 5.0), was the need for a surgical cut-down
for implantation. The Impella CP can be inserted
percutaneously, which enables quick insertion even
JACC VOL. 69, NO. 3, 2017 Ouweneel et al. 285
JANUARY 24, 2017:278–87 Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock

C ENTR AL I LL U STRA T I O N Impella CP Versus IABP in Cardiogenic Shock

Ouweneel, D.M. et al. J Am Coll Cardiol. 2017;69(3):278–87.

(A and B) Schematic drawings of the heart and aorta showing the 2 mechanical support devices used in the study: (A) Impella CP (Abiomed, Danvers,
Massachusetts); (B) the intra-aortic balloon pump (IABP). (C) Time-to-event Kaplan-Meier curves up to 6 months after randomization for all-cause
mortality. LV ¼ left ventricular.
286 Ouweneel et al. JACC VOL. 69, NO. 3, 2017

Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock JANUARY 24, 2017:278–87

before performing primary PCI. Earlier reports have

demonstrated a better survival in patients who REPRINT REQUESTS AND CORRESPONDENCE: Dr.

received a pMCS before primary PCI than in implan-
tation post-PCI (22). Our data also shows a trend to-
ward lower mortality rates in patients in whom either
the device or IABP was initiated before the primary
PCI (25.0% vs. 52.5% overall).
STUDY LIMITATIONS. A major limitation of this trial
is its small number of patients. Adequately powered
randomized clinical trials are needed to ascertain the
value of pMCS in patients with CS after AMI.
CONCLUSIONS
In this explorative study, routine treatment with
pMCS was not associated with lower 30-day mortality
in patients with CS complicating AMI.
ACKNOWLEDGMENTS The authors thank all clinical
and research staff at the participating centers who
made it possible to perform this trial, and thank
Martin Frydland for editing the figures.
José P.S. Henriques, Academic Medical Center, Uni-
versity of Amsterdam, AMC Heart Center, Meiberg-
dreef 9, 1105 AZ Amsterdam, the Netherlands. E-mail:
j.p.henriques@amc.uva.nl.
PERSPECTIVES
COMPETENCY IN PATIENT CARE AND PROCE-
DURAL SKILLS: In patients with AMI complicated by
severe CS, treatment support with a pMCS device was
associated with a 30-day survival rate similar to that
achieved with IABP.
TRANSLATIONAL OUTLOOK: Further studies are
needed to clarify the clinical characteristics of patients
with CS who benefit from 1 mode of mechanical cir-
culatory support more than another.

REFERENCES

1. Hochman JS, Sleeper LA, Webb JG, et al., from
the SHOCK Investigators. Early revascularization in
acute myocardial infarction complicated by
cardiogenic shock. Should we emergently revas-
cularize occluded coronaries for cardiogenic shock.
N Engl J Med 1999;341:625–34.
2. Babaev A, Frederick PD, Pasta DJ, et al. Trends
in management and outcomes of patients with
acute myocardial infarction complicated by
cardiogenic shock. JAMA 2005;294:448–54.
3. Goldberg RJ, Spencer FA, Gore JM, Lessard D,
Yarzebski J. Thirty-year trends (1975 to 2005) in
the magnitude of, management of, and hospital
death rates associated with cardiogenic shock in
patients with acute myocardial infarction: a
population-based perspective. Circulation 2009;
119:1211–9.
4. Thiele H, Zeymer U, Neumann FJ, et al. Intra-
aortic balloon support for myocardial infarction
with cardiogenic shock. N Engl J Med 2012;367:
1287–96.
5. Stretch R, Sauer CM, Yuh DD, Bonde P. National
trends in the utilization of short-term mechanical
circulatory support: incidence, outcomes, and cost
analysis. J Am Coll Cardiol 2014;64:1407–15.
6. Sjauw KD, Engstrom AE, Vis MM, et al.
A systematic review and meta-analysis of intra-
aortic balloon pump therapy in ST-elevation
myocardial infarction: should we change the
guidelines? Eur Heart J 2009;30:459–68.
7. Thiele H, Zeymer U, Neumann FJ, et al. Intra-
aortic balloon counterpulsation in acute myocar-
dial infarction complicated by cardiogenic shock
(IABP-SHOCK II): final 12 month results of a
randomised, open-label trial. Lancet 2013;382:
1638–45.
8. Wijns W, Kolh P, Danchin N, et al. Guidelines on
myocardial revascularization. Eur Heart J 2010;31:
2501–55.
9. Levine GN, Bates ER, Blankenship JC, et al. 2011
ACCF/AHA/SCAI guidelines for percutaneous cor-
onary intervention: a report of the American Col-
lege of Cardiology Foundation/American Heart
Association Task Force on Practice Guidelines and
the Society for Cardiovascular Angiography and
Interventions. J Am Coll Cardiol 2011;58:e44–122.
10. Windecker S, Kolh P, Alfonso F, et al. 2014
ESC/EACTS guidelines on myocardial revasculari-
zation: the Task Force on Myocardial Revasculari-
zation of the European Society of Cardiology (ESC)
and the European Association for Cardio-Thoracic
Surgery (EACTS). Eur Heart J 2014;35:2541–619.
11. O’Gara PT, Kushner FG, Ascheim DD, et al. 2013
ACCF/AHA guidelines for the management of ST-
elevation myocardial infarction: a report of the
American College of Cardiology Foundation/
American Heart Association Task Force on Practice
Guidelines. J Am Coll Cardiol 2013;61:e78–140.
12. Ouweneel DM, Henriques JP. Percutaneous
cardiac support devices for cardiogenic shock:
current indications and recommendations. Heart
2012;98:1246–54.
13. Sjauw KD, Remmelink M, Baan J Jr., et al. Left
ventricular unloading in acute ST-segment eleva-
tion myocardial infarction patients is safe and
feasible and provides acute and sustained left
ventricular recovery. J Am Coll Cardiol 2008;51:
1044–6.
14. Engstrom AE, Cocchieri R, Driessen AH, et al.
The Impella 2.5 and 5.0 devices for ST-elevation
myocardial infarction patients presenting with
severe and profound cardiogenic shock: the Aca-
demic Medical Center intensive care unit experi-
ence. Crit Care Med 2011;39:2072–9.
15. Seyfarth M, Sibbing D, Bauer I, et al.
A randomized clinical trial to evaluate the safety
and efficacy of a percutaneous left ventricular
assist device versus intra-aortic balloon pumping
for treatment of cardiogenic shock caused by
myocardial infarction. J Am Coll Cardiol 2008;52:
1584–8.
16. Ouweneel DM, Engstrom AE, Sjauw KD, et al.
Experience from a randomized controlled trial
with Impella 2.5 versus IABP in STEMI patients
with cardiogenic pre-shock. Lessons learned from
the IMPRESS in STEMI trial. Int J Cardiol 2016;202:
894–6.
17. 18th World Medical Association General As-
sembly. WMA Declaration of Helsinki—ethical
principles for medical research involving human
subjects. Available at: http://www.wma.net/en/
30publications/10policies/b3/. Accessed September
15, 2016.
18. Thiele H, Sick P, Boudriot E, et al. Randomized
comparison of intra-aortic balloon support with a
percutaneous left ventricular assist device in pa-
tients with revascularized acute myocardial
infarction complicated by cardiogenic shock. Eur
Heart J 2005;26:1276–83.
19. Burkhoff D, Cohen H, Brunckhorst C, et al.
A randomized multicenter clinical study to eval-
uate the safety and efficacy of the TandemHeart
percutaneous ventricular assist device versus
JACC VOL. 69, NO. 3, 2017
JANUARY 24, 2017:278–87
conventional therapy with intraaortic balloon
pumping for treatment of cardiogenic shock. Am
Heart J 2006;152:469.e1–8.
20. Manzo-Silberman S, Fichet J, Mathonnet A,
et al. Percutaneous left ventricular assistance in
post cardiac arrest shock: comparison of intra-
aortic blood pump and IMPELLA Recover LP2.5.
Resuscitation 2013;84:609–15.
21. Lauten A, Engstrom AE, Jung C, et al. Percu-
taneous left-ventricular support with the Impella-
2.5-assist device in acute cardiogenic shock:
Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock
results of the Impella-EUROSHOCK-registry. Circ
Heart Fail 2013;6:23–30.
22. O’Neill WW, Schreiber T, Wohns DH, et al. The
current use of Impella 2.5 in acute myocardial
infarction complicated by cardiogenic shock: re-
sults from the USpella Registry. J Interv Cardiol
2014;27:1–11.
23. U.S. Food and Drug Administration. Summary of
safety and effectiveness data (SSED). Available at:
http://www.accessdata.fda.gov/cdrh_docs/pdf14/
P140003S004B.pdf. Accessed September 6, 2016.
Ouweneel et al. 287
KEY WORDS acute myocardial infarction,
cardiogenic shock, intra-aortic balloon pump,
mechanical circulatory support, randomized
controlled trial
A PP END IX For the trial organization and
supplemental methods as well as supplemental
figures and tables, please see the online version
of this article.